Node-level Regression Uncertainty Quantification on Basic
100PICPER-GNN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ER-GNN2026.05 | 100 | 1.24 | |
| BayesianNN2026.05 | 100 | 3.01 | |
| MC Dropout2026.05 | 99 | 0.32 | |
| QpiGNNWidth penalty (lambda)=0.12026.05 | 98 | 0.93 | |
| CF-GNN2026.05 | 92 | 1.9 | |
| RQRadj.-GNN2026.05 | 90 | 0.82 | |
| QpiGNNWidth penalty (lambda)=opt.2026.05 | 90 | 0.3 | |
| QpiGNNWidth penalty (lambda)=0.52026.05 | 89 | 0.3 | |
| SQR-GNN2026.05 | 85 | 0.33 |